A Subtherapeutic International Normalized Ratio Despite Increasing Doses of Warfarin: Could This Be Malabsorption?

A Subtherapeutic International Normalized Ratio Despite Increasing Doses of Warfarin: Could This Be Malabsorption? LUIS F. LARA, MD; LAZARO L. DELGADO, MD; LAWRENCE A. FRAZEE, PHARMD; KATHY M. HAUPT, MD; GREGORY W. RUTECKI, MD

ABSTRACT: Objective: To describe a case of warfarin resistance apparently caused by malabsorption and to review the literature regarding warfarin resistance. Case Summary: A 28-year-old renal transplant patient with systemic lupus erythematosus was admitted for upper extremity thrombophlebitis. Resistance to oral warfarin was demonstrated. Potential causes were investigated. The trapezoidal rule was used to compare the area under the curve for intravenous versus oral dosing of

warfarin. The usual bioavailability of warfarin should be 100%. In this patient, warfarin bioavailability after oral dosing was 1.5%. Three potential causes, malabsorption (FF), enzymatic degradation (FG), and first-pass extraction in the portal circulation (FH), are discussed. Conclusion: This case demonstrates resistance to warfarin presumably caused by malabsorption. KEY INDEXING TERMS: Warfarin malabsorption; Warfarin resistance.

W

presumed because the defect was demonstrated in appropriate family members. Interestingly, because the absorption and clearance of warfarin were normal, the mechanism of resistance probably occurred at a molecular level because of an enhanced formation of reduced vitamin K or resistance of vitamin K reductases to the action of warfarin. 3--5,8-12 Acquired resistance may result from noncompliance, excessive vitamin K ingestion, or drugs that affect the pharmacokinetics of warfarin through inhibition of absorption or enhancement of biotransformation (Table 1).3-6,13 We report a case ofwarfarin resistance presumed secondary to malabsorption of orally administered warfarin.

arfarin is an oral anticoagulant that exerts its pharmacological action by inhibiting the activation of the vitamin K-dependent clotting factors II, VII, IX, and X. Normally, reduced vitamin K is oxidized to vitamin K 2,3-epoxide as a cofactor in the 'Y-carboxylation of these clotting factors. Warfarin acts on the vitamin K reductases, thereby inhibiting the reduction of vitamin K 2,3-epoxide back to vitamin K. The end result is therapeutic anticoagulation primarily through an effect on the extrinsic clotting system measured by the international normalized ratio (INR).l Warfarin is well absorbed from the gastrointestinal tract after oral or rectal dosing with an absolute bioavailability of up to 100%. Its onset and duration of action are also predictable. 1- 4 Resistance to the effects of warfarin is uncommon and may be either hereditary or acquired. 1 ,2,5-7 Hereditary resistance is rare, with four previous cases confirmed. Autosomal dominant inheritance is

From the Department of Internal Medicine, Northeastern Ohio Uniuersities College of Medicine (NEOUCOM) Affiliated Hospitals at Canton, Ohio (LFL, LLD); Department of Internal Medicine, NEOUCOM, Akron General Medical Center, Akron, Ohio (LAF); Department of Family Practice, NEOUCOM, Aultman Hospital, Canton, Ohio (KMH); and Internal Medicine Residency Program, Euanston Northwestern Healthcare (GWR). Submitted: Occtober 27, 1999; accepted in reuised form March 2,2000. Correspondence: Gregory W. Rutecki, M.D., Euanston Northwestern Healthcare, Department of Internal Medicine, 2650 Ridge Aue., Euanston, IL 60201.

214

[Am J Med Sci 2000;320(3):214-8.]

Case Report A 28-year-old female with systemic lupus erythematosus, who had undergone kidney transplantation from a cadaveric donor 4 years before, was admitted for swelling of her left arm. A Hickman catheter had been placed 30 days before admission in her left subclavian vein for venous access. Her oral medications included prednisone (10 mg per day), cyclosporine (100 mg twice per day), azathioprine (125 mg per day), trimethoprim-sulfamethoxazole (160/800 mg per day), calcium (1200 mg per day) and vitamin D (1 /log per day), clonidine (0.2 mg per day), valproic acid (250 mg three times per day), omeprazole (20 mg twice per day), an albuterol inhaler [2 puffs (180 /log) every 6 hours), promethazine (25 mg every 6 hours), and nefazodone (50 mg in the morning, 100 mg at noon, and 150 mg at bedtime). Her left arm was edematous, erythematous, and tender. A deep venous thrombosis involving the subclavian, axillary, and proximal brachial veins was demonstrated by Doppler examination. The patient was admitted to the hospital for anticoagulation and removal of the Hickman catheter. Parenteral anticoagulation was begun with enoxaparin 1 September 2000 Volume 320 Number 3

Lara et al

Table 1. Acquired Resistance to Warfarin Acquired Noncompliance Drugs that affect pharmacokinetics by - Inhibiting absorption: cholestyramine - Enhancing biotransformation: alcohol, haloperidol, barbiturates, griseofulvin, rifampin, nafcillin, meprobamate, carbamazepine Enhanced formation of clotting factors due to oral contraceptives, adrenocortical steroids, pregnancy, ingestion or administration of vitamin K and azathioprine? Hypoproteinemia (increases volume of distribution and shortens half life) Decreased drug absorption caused by short bowel syndrome or selective malabsorption. Hypothyroidism Hereditary Tissue resistance caused by: - Enhanced reduction of vitamin K - Defects in the microsomal vitamin K reductases - Resistance of the vitamin K reductases to warfarin

mg/kg subcutaneously every 12 hours and warfarin was started 12 hours later at 5 mg a day. Workup for hypercoagulability demonstrated heterozygosity for the Factor V Leiden mutation as well as the presence of the lupus anticoagulant (SmithKline Beecham Laboratory, Valleyview, OH). Despite escalating doses of warfarin, up to 50 mg orally, the INR remained subtherapeutic. Nursing personnel supervised the ingestion of warfarin and the dose was administered in variable relationship to other medications to exclude drug interference with warfarin absorption. Diet was reviewed to rule out excess vitamin K. Random blood samples were drawn to measure warfarin levels, which were low to undetectable. The suspicion of warfarin malabsorption prompted further investigation.

Methods Study Design. A 3-phase protocol was developed to use the trapezoidal rule for determination of the area under the curve (AUC) comparing serum warfarin levels after an intravenous versus oral challenge. 9 The patient's bioavailability after oral dosing could then be calculated. The Institutional Review Board approved the protocol and the patient signed an informed consent to participate in the study. Laboratory Measurements of Serum Warfarin Levels. An internal standard (8-methoxy-psoralen) was added to 0.5 mL aliquots that were diluted with 2.5 mL of deionized water and made acidic with 0.2 mL of 1.0 M acetic acid. Samples were extracted with a solid phase extraction procedure using International Sorbent Technology Isolute HCX extraction column on the Zymark Rapid Trace automated extraction instrument. The extraction columns were conditioned with methanol, water, and 1.0 M acetic acid. The pH-adjusted samples were then added to the column and the bed was washed with a pH 6.0 phosphate buffer, 1.0 M acetic acid, and hexane. Warfarin and the internal standard were eluted with methylene chloride, which was evaporated to dryness and reconstituted with 200 /-LL of mobile phase. Twenty microliters of the mobile phase was injected onto the HPLC column. HPLC analysis was on a Syncropac SCD 100, 4.6 x 250 mm with an Upchurch Scientific type "R" precolumn. The mobile phase (1.5 mUmin) consisted of 250 mL of acetonitrile and 750 mL of 0.1 N phosphoric acid that had been adjusted to a pH of5.7 using triethylamine. Initially, an UV detector set of 276 nm was used, but to detect below the usual limit of quantification, a fluorescence detector THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES

Table 2. Intravenous Dose, 20 mg Time (hours)

o

0.5 1.5 2.75 6.5 10.25 16.5 22.5

Warfarin level (/-Lg/mL) 0 2.3 2.5

1.9

1.9

1.4

1.6

1.3

was used at excitation and emission settings of 290 and 385 nm, respectively. The calibrators for the UV detection analysis ranged from 0.20 to 10 /-Lg/mL, and the calibrators used for the fluorescence detection analysis ranged from 0.01 to 10 /-Lg/mL. Warfarin and the internal standard eluted at 6.12 and 8.52 minutes, respectively. The internal standard was not detectable at the excitation and emission settings used; therefore, the absolute response was determined by the UV detector. The limits of quantification for the UV and fluorescence methods were 0.20 and 0.01 /-Lg/mL, respectively. Phase I. After admission to the hospital, baseline activated partial thromboplastin time (aPTT), prothrombin time (PT), and vitamin K levels were obtained and were normal. The patient then received an intravenous dose of 20 mg of warfarin (Coumadin; DuPont, Wilmington, DE). Serial serum warfarin measurements were obtained after administration of the drug over the next 24 hours (Table 2). Phase II. After a washout period of 7 days, the patient was readmitted to the hospital. Baseline aPTT, PT, and vitamin K levels were normal. The patient received an oral dose of 60 mg of warfarin (6 10-mg tablets of Coumadin, DuPont, Wilmington, DE). Serial warfarin levels were obtained for the next 24 hours. Serum levels of warfarin were undetectable throughout the entire sampling period making the calculation of the bioavailability impossible. Phase III. Twenty-one days after phase II, the patient was readmitted to the hospital. Baseline aPTT, PT, and vitamin K levels were normal. The patient received an oral dose of 120 mg of warfarin. Using a serum warfarin level detection limit of 0.20 /-Lg/mL, a level of 0.26 /-Lg/mL was measured 0.5 hours after dosing. No other levels were detected and the tests were repeated. This time, the test detection limit was decreased to 0.01 /-Lg/mL by fluorescence method. This identified levels over a 24-hour period (Table 3). The patient refused a 72-hour stool collection after warfarin loading.

Results Data Analysis. The bioavailability for the orally administered warfarin was determined by calculating the AUC using the trapezoidal rule (Table 4). The bioavailability is depicted in graph form in Figure 1. Warfarin has an absolute bioavailability of 100%. The bioavailability of warfarin in this patient was 1.5% (Table 5). Serum cyclosporine and vitamin B 12 levels were therapeutic. Thyrotropin-stimulating hormone was normal.

Table 3. Oral Dose, 120 mg Time (hours)

o

0.5

4.5

5

7.5

Warfarin level (/-Lg/mL) 0 0.26 0.04 0.04 0.03

12

23

0.03

0.03

215

Warfarin Resistance Apparently Caused by Malabsorption

Table 4. AUC and Bioavailability

Table 5. Comparison of Oral and Intravenous Warfarin Doses by AUC and Bioavailability

Area Under the Plasma Concentration-Time Curve (AUC) AUC = [(Co + C Ctl)) X (tl) - 0)] + [(CCU) + CCt2») + [(CCtn-l) + CCtn») X (tn - (tn - l})] + [CCtl/ke ke Co tn tp CCtn) Cp ke

Oral X

Intravenous

(t2 - tl)] Dose (mg) AUC CO--4) (mg X hr/mL) Bioavailability

= [lnC p - 1nCCtn)]I[tn-tp] = Plasma concentration at baseline

the nth time measurement time measurement of the peak plasma concentration = plasma concentration at the nth time measurement = peak plasma concentration = elimination rate constant =

a

120

20

9.55 0.015a

104.81 1.000

Therefore, the absolute oral bioauailability is l.5%.

=

Bioavailability (F) F

=

[AUC

PO

X

dose rv]/[AUC rv

X

dosepo]

AUCpo = area under the plasma concentration-time curve after the oral dose AUC rv = area under the plasma concentration-time curve after the intravenous dose

Discussion

The absolute systemic bioavailability of an orally administered drug is dependent on 3 general factors. The fraction of the dose that enters gastrointestinal

tissues from the lumen and escapes decomposition or loss in the feces is often referred to as FF. Once the drug is present in gut tissue, degradation via microsomal enzymes present in the gut wall may result in further loss of the administered drug. The fraction that escapes metabolism in the gut wall is referred to as FG. Finally, if the drug is metabolized in the liver, it may be extracted presystemically during its first pass through the portal circulation. The fraction that escapes first pass metabolism is known as FR. Therefore, the overall systemic bioavailability (F) of an orally administered drug can be described by the equation F = FF X FG X FR. For example, if the extraction of a given drug at each of these steps is 10%, then the absolute sys-

---_._._------_ .._----------_.-- - -

3

2.5 ~

~Ol u

g

2

c 0

-

~ 'c Q) u

c 0

()

c ·c

{g

~ 0.5

o

2

3

4

5

6

7

8

9

10 11

12 13 14 15 16 17 18 19 20 21

Hours after Dose

22 23 24

.20 mg. I.V. Warfarin 0120 mg. P.O. Warfarin

Figure 1. Comparisons of the serum concentrations of oral versus intravenous warfarin over a 24-hour period.

216

September 2000 Volume 320 Number 3

Lara et al

temic bioavailability of that drug will be F = 0.9 X 0.9 X 0.9 = 0.729, demonstrating that approximately 73% of the administered dose reaches the systemic circulation. An orally administered drug can be unavailable at any of the three steps between the lumen of the gut and the systemic circulation. 14 Normally, warfarin has an absolute bioavailability of 100%. Therefore, the loss of warfarin in the feces, degradation in the gut wall, or first-pass metabolism of warfarin in the liver is presumed to be negligible. Inhibition or induction of extraction at any of these steps could affect the bioavailability of warfarin. Decreased bioavailability of warfarin has been described twice previously in the literature. The first report was of a woman who had been therapeutically anticoagulated for 2 years on oral warfarin and then became resistant. An analysis of bioavailability using a 30-mg intravenous and oral dose of warfarin was performed. Plasma warfarin samples were collected after both doses and the AUe was calculated. The bioavailability was found to be 33%. Administration of another hydroxycoumarin derivative, dicumarol, also failed to produce anticoagulation. Phenindione, an indandione derivative, was able to elicit a satisfactory anticoagulant response. This was the first report of malabsorption of coumarin-type anticoagulants. 9 A second case of warfarin malabsorption was recently described in a patient with short bowel syndrome secondary to resection (duodenectomy, gastrojejunostomy) after multiple gunshot wounds to the abdomen. 15 The patient received escalating doses of warfarin, up to 20 mg a day, with only minimal increase in the INR. The authors concluded that malabsorption because of short bowel syndrome was the cause of warfarin resistance. Plasma warfarin concentrations were not measured to assess bioavailability. In addition, stool was not collected. Drugs may also affect warfarin pharmacokinetics by either decreasing absorption from the gastrointestinal tract or by changing warfarin's metabolic clearance.1,3,6,16,17 For example, cholestyramine impairs warfarin absorption through binding. In addition, barbiturates, rifampin, and carbamazepine, through an induction of hepatic oxidases, may decrease the effect of warfarin. There is no evidence in the literature for nefazodone interactions with medications the patient was taking. In a review by Wells et al,16 azathioprine and cyclosporine have level 3 evidence for interaction with warfarin. Level 3 evidence is documented by case reports with plausible timing. Our patient was on both of these drugs. A MEDLINE search back to 1966 reveals two reports of azathioprine interacting with warfarin 18 ,19 and a report involving 6-mercaptopurine. 20 These reports demonstrated an unusually high warfarin requirement (15-20 mg daily) with concurrent azathioprine or 6-mercaptopurine administration. A possible drug interaction was postulated. UnfortuTHE AMERICAN JOURNAL OF THE MEDICAL SCIENCES

nately, these reports failed to document the mechanism of interaction and no attempt was made to determine the AUe to possibly explain malabsorption as the cause of warfarin resistance. One study from 1977,21 of rats, determined that 6-mercaptopurine shortened prothrombin time because of an increase in factor II activity. If this is indeed the mechanism, warfarin plasma levels would not be affected by either 6-mercaptopurine or azathioprine. Therefore, our patient's warfarin resistance is not likely to have been caused by these drugs because, in our case, decreased bioavailability was clearly demonstrated, suggesting malabsorption. There is minimal data on a possible interaction between warfarin and cyclosporine. A MEDLINE literature search back to 1966 identified one related letter published in 1988. 22 The addition of warfarin therapy to a patient receiving cyclosporine resulted in decreased cyclosporine levels. When the dose of cyclosporine was increased, prothrombin activity increased from 17% of control to 64%, requiring an increase in warfarin dosage. In addition, the patient in that case report was also taking phenobarbital, which can decrease concentrations of both warfarin and cyclosporine. In this case, the suspected mechanism of resistance to warfarin did not involve malabsorption, but rather cyclosporine-warfarin interactions leading to decreased drug availability. Based on the evaluation conducted in our patient, this cyclosporine-warfarin interaction could not account for the almost undetectable serum levels we obtained. Conclusion

The patient reported herein has resistance to warfarin that was presumably caused by malabsorption. Although liver first-pass clearance of warfarin (FF) cannot be completely ruled out, this mechanism seems less likely. Unfortunately, because the patient refused a stool sample to assay unabsorbed warfarin, the mechanism of warfarin resistance cannot be completely elucidated. The patient is presently being treated successfully with daily doses of subcutaneous enoxaparin. We thank National Medical Services, Inc. (Philadelphia, PA) for their careful measurement of serum warfarin levels and Norman Sohar, R.N., for help in the preparation of this report. References 1. Majerus P, Broze GJ, Miletich JP, et al. Anticoagulant,

thrombolytic, and antiplatelet drugs. In: Hardman JG, Limbird LE, et aI, editors. Goodman & Gilman's the pharmacological basis of therapeutics. 9th ed. New York: McGraw Hill; 1995. p. 1346-50. 2. Diab F, Feffer S. Hereditary warfarin resistance. South Med J 1994;87:407-9.

217

Warfarin Resistance Apparently Caused by Malabsorption

3. Hallak HO, Wedlund PJ, Modi MW, et al. High clearance of (s)-warfarin in a warfarin resistant subject. Br J Clin Pharmacol 1993;35:327-30. 4. Kereveur A, LecIerq M, Trossaert M, et al. Vitamin K metabolism in a patient resistant to vitamin K antagonists. Haemostasis 1997;27:168-73. 5. Talstad I, Gamst ON. Warfarin resistance due to malabsorption. J Intern Med 1994;236:465-7. 6. Sanchez Villegas JM, Piqueras J, Nicolau I, et al. Resistancia a los antivitaminicos K. Sangre 1992;37:297-8. 7. Warrier I, Brennan CA, Lusher JM. Familial warfarin resistance in a black child. Am J Pediatr Hematol Oncol 1986;8:346-57. 8. Fraser GL, Miller M, Kane K. Warfarin resistance associated with nafcillin therapy. Am J Med 1989;87:237-8. 9. Misenheimer TM, Lund M, Miller-Baker AE, et al. Biochemical basis of warfarin and bromadiolone resistance in the house mouse, Mus musculus domesticus. Biochem Pharmacol 1994;47:673-8. 10. Hulse ML. Warfarin resistance: diagnosis and therapeutic alternatives. Pharmacotherapy 1996;12:1009-17. 11. Lefrere JJ, Horellou MH, Conard J, et al. Proposed classification of resistances to oral anti-coagulant therapy. J Clin Pathol 1987;40:242. 12. Vermeer C, Soute BAM, Aalten M, et al. Vitamin K reductases in normal and in warfarin resistant rats. Biochem Pharmacol 1988;14:2876-8.

218

13. Absorption. In: Rowland M, Tozer TN, editors. Clinical pharmacokinetics: concepts and applications. Philadelphia: Lea & Febiger; 1989. p. 113-30. 14. Brophy DF, Ford SL, Crouch MA. Warfarin resistance in a patient with short bowel syndrome. Pharmacotherapy 1998; 18:646 -9. 15. Wells PS, Holbrook AM, Crowther NR, et al. Interaction of warfarin with drugs and food. Ann Intern Med 1994;121: 676-83. 16. Michalets EL. Update: clinically significant cytochrome P-450 drug interactions. Pharmacotherapy 1998;18:84-112. 17. Hirsch J, Dalen JE, Deykin D, et al. Oral anticoagulants: mechanism of action, clinical effectiveness, and optimal therapeutic range. Chest 1995;108(Suppl):231S-46S. 18. Singleton JD, Conyers L. Warfarin and azathioprine: an important drug interaction [letter] Am J Med 1992;92:217. 19. Rivier G, Khamashta MA, Hughes GR. Warfarin and azathioprine: a drug interaction does exist. Am J Med 1993; 95:342. 20. Spiers AS, Mibashan RS. Increased warfarin requirement during mercaptopurine therapy: a new drug interaction [letter]. Lancet 1974;2(7874):221-2. 21. Martini A. Jahnchen E. Studies in rats on the mechanism by which 6-mercaptopurine inhibits the anticoagulant effect of warfarin. J Pharmacol Exp Ther 1977;201:547-53. 22. Snyder DS. Interaction between cyclosporine and warfarin [letter]. Ann Intern Med 1988;108:311.

September 2000 Volume 320 Number 3